We propose a joint embedding space populated by both shapes and images, where the Euclidean distance between embedded entities reflects the similarity between the objects contained or represented by their original form. This embedding space then facilitates comparison between entities of any form, and allows for cross-entity-based retrieval.

Yangyan Li, Stanford University
Daniel Cohen-Or, Tel Aviv University
Hao Su, Stanford University
Charles Qi, Stanford University
Leonidas Guibas, Stanford University
Noa Fish, Tel Aviv University